Semi-Supervised Learning

Semi-Supervised Learning

1. Choosing Good Unlabeled data

   • Sampling Function
     • IRS
       • Increase the size of S
       • X
     • SRS
       • SB criterion
         • pi, qi
       • Modified criterion
         • pi, qi and Pe
     • Hybrid = IRS +SRS
       • X
   • Multi-view technique
     • Feature Selection
       • Individual Feature Selection
     • View decomposition
       • Genetic
       • Cuckoo
       • Entropy
         • X
       • Information Theory

2. Classifier

   • SVM
     • TSVM
     • S3VM
     • S3VM-us
     • LapSVM
   • Boosting
     • H <- H + a*h
       • Adaboost
       • Semiboost
       • Marginboost
       • SSMB
       • SSMB2
   • 3DS and 4DS
     • X

3. Clustering --> Sampling

   • X